Avi Cluster "Uptime" is not accurate when Avi Cluster status changes ( Rebooted/Added/Updated ) triggering AKO pod reboot(s)

Products

• VMware Avi Load Balancer

Issue/Introduction

AKO pod restarts as the uptime of the Avi cluster nodes changes when the follower node re-joins the cluster. 

 

Cause

The time was reset for all the nodes, although only the follower node is joining the cluster.

 

 

Resolution

The issue is fixed in 30.2.1

Additional Information

To view the uptime for each node of the cluster, the below command can be used from Avi command line (bash prompt) 
# uptime

To view the uptime of all nodes of the cluster, below command is used from Avi Cli prompt
> show cluster nodes

Below is a sample output for the above command. The point to consider in the below command is "up since" timestamp for all nodes.  As a result of this bug, the "up since" timestamp is changing for all three nodes of the cluster , even if only one of the follower nodes is joining the cluster (due to reboot or a new node being added). 

The impact can be observed: 

• This can lead to misinformation about cluster uptime when using the command "show cluster nodes" on Avi CLI. 
• If AKO is used in the environment, it can cause AKO pod restarts.